Logistics resilience is often discussed as though it can be bought at short notice: add another carrier, increase safety stock, lease extra warehouse space or activate an alternative route. Those measures have value, but they do not solve the harder problem. When infrastructure is fixed, slow to expand or difficult to substitute, the operating model may remain fragile even when contingency plans look comprehensive on paper.
Recent uncertainty around shipping through the Strait of Hormuz, delays surrounding Heathrow’s potential third runway and the continuing debate over warehouse automation all point to the same strategic issue. Capacity is only resilient when it is usable, accessible and replaceable. A second theoretical option is not a genuine alternative if it depends on the same congested gateway, requires unavailable handling equipment or cannot support the required service level.
For logistics leaders, the priority should therefore move beyond counting suppliers and routes. The more important question is: how much operational choice has the network actually preserved?
Map the infrastructure dependencies that cannot be replaced quickly
Most organisations have a reasonable view of their direct logistics suppliers. Fewer have a detailed understanding of the infrastructure beneath those relationships. A carrier may be diversified, but its services could still rely on the same port, airport, rail terminal, road corridor, border crossing or energy supply.
This creates hidden concentration risk. Several apparently independent transport options may converge on one physical bottleneck. The same is true inside the warehouse. Multiple customer channels may depend on one automated storage system, one controls platform, one specialist maintenance team or one building with limited expansion potential.
A useful resilience review should identify these dependencies at asset level. For each critical flow, businesses should record the infrastructure involved, the time required to replace it, the practical alternatives available and the conditions under which those alternatives would fail. The objective is not to eliminate every dependency. That would be unrealistic and expensive. It is to distinguish manageable exposure from exposure that cannot be mitigated within the required recovery window.
Plan for capacity that is genuinely usable
Disruption planning often overstates the value of nominal capacity. A different port may have available slots, but not the right equipment. An alternative airport may have runway capacity but lack the handling, customs or road connections needed for a particular product. A secondary warehouse may have floor space but no labour pool, charging infrastructure or systems integration.
Capacity planning must therefore include an operational qualification process. That means testing whether an alternative site or mode can handle the actual product range, volumes, cut-off times, documentation requirements and peak conditions. It also means establishing commercial access before disruption occurs. A route that only becomes available after a crisis begins may be too late, too expensive or too congested to support the business.
Multimodal planning should be treated in the same way. Road, rail, short-sea and air options should not be listed merely to demonstrate diversity. They should be assessed against specific disruption scenarios, including lead times for switching, equipment availability, terminal capacity, border requirements and the knock-on effects for inventory and customer commitments.
Clear triggers are essential. If a chokepoint is disrupted for three days, what changes? At seven days, does the business switch mode, ration capacity or reallocate stock? At fourteen days, which customers or product lines receive priority? Decisions made in advance are more reliable than emergency choices made when every operator is competing for the same scarce alternatives.
Do not let automation create a new single point of failure
Warehouse automation can improve throughput, safety and labour productivity, but it can also make flexibility harder if the design is too tightly matched to one forecast or operating pattern. The key question is not simply whether a building is large enough for automation. It is whether the system can adapt when volumes, product dimensions, order profiles, labour availability or network flows change.
Before committing to a highly integrated solution, operators should examine the consequences of partial failure. Can manual processes continue at a reduced rate? Can equipment be bypassed? Are critical spare parts available locally? Is there more than one qualified maintenance provider? Can software changes be made without lengthy re-engineering? What happens if a customer’s profile changes faster than the automation’s payback model assumed?
Modularity is not automatically the answer, and manual operations are not automatically more resilient. The right design depends on the required service level and the cost of interruption. However, a system that delivers maximum efficiency only under a narrow set of conditions may be less valuable than one that can operate at 70 or 80 per cent capacity across a wider range of scenarios.
That trade-off should be made explicitly. Resilience has a cost, but so does inflexibility. A lower headline cost can conceal greater exposure to downtime, specialist support shortages and expensive redesign.
Build flexibility before the disruption
Infrastructure decisions are difficult to reverse. New airports, terminals, roads, rail connections and warehouses take years to plan and deliver. Automation investments can lock in layouts, processes and technology choices for a decade or more. By the time a weakness becomes visible in operational performance, the window for a low-cost correction may have closed.
Boards and supply-chain leaders should therefore treat flexibility as a design requirement rather than an emergency response. This could mean reserving expansion space, maintaining compatible manual capability, qualifying secondary sites, negotiating priority access to alternative modes or staging automation in smaller increments. It may also mean accepting some underused capacity as an insurance premium against severe disruption.
The strongest networks will not necessarily be those with the most assets. They will be those with the greatest ability to switch between assets, modes and operating methods without losing control of service, cost or compliance. That requires investment, but it also requires better questions during planning: which choices will remain open, which dependencies will become permanent and how quickly can the business change course?
Geopolitical chokepoints, delayed infrastructure projects and ambitious automation programmes are different issues on the surface. Operationally, they expose the same weakness: a logistics model that has optimised around capacity it cannot easily replace. Resilience begins when businesses identify those constraints early, define the triggers for action and build credible alternatives before the network is under pressure.
